[LEAPSECS] civil-solar correlation with TI

Zefram zefram at fysh.org
Tue Dec 23 11:33:28 EST 2008

Daniel R. Tobias wrote:

>And then your distant descendants will throw a huge fit about the

>possibility of a leap hour being imposed, which would be much more

>noticeably disruptive than leap seconds, and will indefinitely block

>such a thing, until some day noon comes in the middle of the night.

I, for one, don't see a problem with the time of solar midnight being
labelled "12:00". If office hours are consistently 21:00 to 05:00, and
particularly if they've been something like that for your whole life,
I think you'd cope. But that's not what I want to talk about here.

Suppose that people in the future overwhelmingly want local civil time
of day to continue to approximate local solar time of day. Suppose also
that since 2020 civil time has been based on TI, defined as TAI - 38 s,
instead of being based on UTC. (Whether the UTC service continues to
exist is irrelevant here, as is the possibility of TI being misleadingly
named "UTC".) What do people actually do? Presumably they don't let
things get as far as solar midnight being at 12:00.

I think in this situation it is inevitable that the timezones will shift.
We already have political debates about changing timezone offsets to have
some target relationship to daylight hours. This debate is partly about
twice-annual offset jumps, but also partly about permanent changes to the
base offset. The very same process, if continued in this hypothetical
situation, will result in offsets gradually jumping downwards over a
period of millennia. After two millennia we might expect Britain's
timezone to be about TI-3h, and the west coast of North America to be
about TI-11h.

There are then two ways that this can continue. The first way is that
offsets just continue their gradual change, with no one doing anything
more radical. I think this is a reasonable default prediction when
forecasting politics. By the year 5000 no one uses a positive offset
from TI. After five millenia Britain might be on TI-65h. The local
date gets to be substantially different from the TI date, to the same
sort of degree that the local hour-of-day currently differs from the UTC
hour-of-day. Timezone offsets continue to grow secularly without bound.
Sounds unworkable? Possibly, but I can't put my finger on any reason
why large offsets per se would cause the system to collapse.

The other option is that people also find large offsets from TI
sufficiently odious to do something about. What they do is to jump
across the International Date Line. Countries already do this from time
to time. The IDL has already, in little over a century, drifted a couple
of timezones eastwards from its nominal position. Kiritimati is now on
UT+14h, whereas the most negative current offsets that I see in Olson
are UT-11h. So in this hypothetical situation, when a country has got
as far as TI-13h or so it'll skip a calendar date and jump to TI+11h.
The IDL will secularly drift eastwards, repeatedly wrapping right around
the globe. This scheme is exactly as workable as the IDL itself, which is
an inherent feature of the civil-solar correlation that we're presuming.

If you're worried about the violence being done to the calendar in
either of these situations, bear in mind that after a few millenia
the Gregorian calendar will desynchronise from the seasons anyway.
This problem exists regardless of what we do to time of day.

Either of my scenarios still suffers from the problem that the TI-UT
difference accelerates. These timezone offset changes would be needed
at decreasing intervals. By the time timezones are jumping by an hour
every year, one might expect to see political pressure for a new scheme.
But I don't think anything would put a stop to the process before then.

Secular growth of timezone offsets has previously been mentioned here,
as one of the unthinkable outcomes. But I think it's a lot more feasible
than it's been given credit for. I've shown two viable scenarios using
only TI and existing timezone mechanisms: there is no requirement for
new technical systems, nor for deliberate international cooperation,
nor even particularly explicit public awareness of the issues. This is
a likely outcome of adopting TI without any further planning.


More information about the LEAPSECS mailing list